TRES
TRiple Evolution Simulation package
Install / Use
/learn @amusecode/TRESREADME
TRES
TRiple Evolution Simulator
Description
TRES is a numerical framework for simulating hierarchical triple systems with stellar and planetary components. Mass transfer from one star to another and the consequential effect to the orbital dynamics is realized via heuristic recipes. These recipes are combined with three-body dynamics and stellar evolution inluding their mutual influences.
TRES includes the effects of common-envelope evolution, circularized stable mass transfer, tides, gravitational wave emission and up-to-date stellar evolution including chemically homogeneous evolution. By default the stellar evolution code SeBa is used. Other stellar evolution codes such as SSE or MESA can also be used.
This document contains the following parts:
TRES-Exo for exoplanet research
TRES with other stellar evolution codes
Compilation
TRES makes use of the Astrophysical Multipurpose Software Environment (AMUSE) See https://amusecode.github.io/ for how to install AMUSE. Note that for standard TRES usage, the only necessary community code to install is SeBa.
Thus, after installing the AMUSE pre-requisites, we can simply install the minimal framework and then add SeBa:
pip install [--user] amuse-framework
pip install [--user] amuse-<seba>
After compiling AMUSE, TRES needs to be installed and compiled by means of the Makefile as following:
First, clone the TRES github repository:
git clone https://github.com/amusecode/TRES.git
Then, from the root of the cloned respository compile the Makefile:
cd seculartriple_TPS
make clean
make
Note: If the above line doesn't work, which may be the case for older versions of amuse, try:
(cd seculartriple_TPS)
mv Makefile Makefile_new
mv Makefile_old Makefile
make clean
make
Simple examples of runs
To evolve a single system with the parameters: primary mass M=1.2 Solar mass, secondary mass m=0.5 Solar mass, tertiary mass m=0.6 Solar mass, inner and outer eccentricity e=0.1 & 0.5, inner and outer orbital separation a=200 & 20000 Solar radii, metallicity z=0.001, and time T=10 Myrs, you need to run:
python TRES.py -M 1.2 -m 0.5 -l 0.6 -E 0.1 -e 0.5 -A 200 -a 20000 -z 0.001 -T 10
assuming AMUSE is loaded in your python.
To include eccentric mass transfer, add '--no_stop_at_mass_transfer' flag, e.g.
python TRES.py -M 1.2 -m 0.5 -l 0.6 -E 0.1 -e 0.5 -A 200 -a 20000 -z 0.001 -T 10000 --no_stop_at_mass_transfer
Input parameters
The full list of possible input parameters is: Depreciated (yet still functioning) parameters are given in {}.
parameter unit / default
--M1 {-M} inner_primary_mass in Solar mass
--M2 {-m} inner secondary mass in Solar mass
--M3 {-l} outer mass in Solar mass
--Ain {-A} inner semi major axis in Solar radius
--Aout {-a} outer semi major axis in Solar radius
--Ein {-E} inner eccentricity
--Eout {-e} outer eccentricity
-i, -I relative inclination in rad
--Gin {-G} inner argument of pericenter in rad
--Gout {-g} outer argument of pericenter in rad
--Oin {-O} inner longitude of ascending node in rad
(outer longitude of ascending nodes = inner - pi)
-Z {-z} metallicity default = 0.02 (Solar)
-t, -T end time in Myr
-N, -n integer number asigned to the triple default = 0
-s integer number to specify seed default = -1 (random seed)
-f name of output file default = TRES
-F type of output file (hdf5/txt) default = hdf5
--dir_plots directory for plots for debugging default = "" (current directory)
mode (aka REPORT_DEBUG == True)
--CE which type of modelling for common envelope evolution default = 2
options:
0: alpha-ce + alpha-dce
1: gamma-ce + alpha-dce
2: seba style; combination of gamma-ce, alpha-ce & alpha-dce
Note that in all cases a double common-envelope is calculated using the alpha-ce.
--SN_kick_distr supernova kick distribution default = 10
options:
0: No kick
1: Hobbs, Lorimer, Lyne & Kramer 2005, 360, 974
2: Hobbs, Lorimer, Lyne & Kramer 2005, 360, 974 scaled down for bh by mass
3: Arzoumanian ea 2002, 568, 289
4: Arzoumanian ea 2002, 568, 289 scaled down for bh by mass
5: Hansen & Phinney 1997, 291, 569
6: Hansen & Phinney 1997, 291, 569 scaled down for bh by mass
7: Paczynski 1990, 348, 485
8: Paczynski 1990, 348, 485 scaled down for bh by mass
9: Verbunt, Igoshev & Cator, 2017, 608, 57
10: Verbunt, Igoshev & Cator, 2017, 608, 57 scaled down for bh by mass
--max_CPU_time maximum CPU time allowed (only works in combination with "stop_at_CPU_time")
default = 3600 (seconds)
Additionally, there is a list of stopping conditions that determines whether the simulation of a system should stop at a certain evolutionary phase. By default, these stopping conditions are set to True, which means they are in effect. However, the four specific mass transfer cases (stable, unstable, eccentric stable & eccentric unstable) are set to False by default. Once "--no_stop_at_mass_transfer" is set to False, it is possible to set the specific mass transfer cases to True.
action items add these to:
--no_stop_at_mass_transfer avoid stopping the simulation at the onset of mass transfer
--no_stop_at_init_mass_transfer avoid stopping the simulation if there is mass transfer initially
--no_stop_at_outer_mass_transfer avoid stopping the simulation when tertiary initiates mass transfer
methodology is as of yet non-existent
--stop_at_stable_mass_transfer stop the simulation at the onset of stable mass transfer in the inner binary
--stop_at_unstable_mass_transfer stop the simulation at the onset of unstable mass transfer in the inner binary (leading to common-envelope evolution)
--stop_at_eccentric_stable_mass_transfer stop the simulation at the onset of stable mass transfer in the inner binary if the orbit is eccentric
--stop_at_eccentric_unstable_mass_transfer stop the simulation at the onset of unstable mass transfer in the inner binary if the orbit is eccentric
--no_stop_at_merger avoid stopping the simulation after a merger
--no_stop_at_inner_collision avoid stopping the simulation after a collision in the inner binary
--no_stop_at_outer_collision avoid stopping the simulation after a collision involving the outer star
--no_stop_at_disintegrated avoid stopping after the system disintegrated into seperate systems
--stop_at_semisecular_regime stop the simulation if the sytem is in the semi secular regime
--stop_at_SN stop the simulation when a supernova occurs
--stop_at_CPU_time stop the simulation when the computational time exceeds a given value
Multiple systems with specified parameters
If you need to follow the triple evolution for multiple systems with parameters which are already specified you can start TRES multiple times, e.g.
python TRES.py -M 1.2 -m 0.5 -l 0.6 -E 0.1 -e 0.5 -A 200 -a 20000 -z 0.001 -T 10
python TRES.py -M 1.5 -m 1 -l 0.6 -E 0.1 -e 0.5 -A 50 -a 20000 -z 0.001 -T 10
python TRES.py -M 1.5 -m 1 -l 0.05 -E 0.1 -e 0.5 -A 50 -a 20000 -z 0.02 -T 10
This is probably not handy for more than 5 systems. Although this can be added in e.g. a shell or Python script.
Random population
A random population can be generated with TPS.py with a Monte Carlo based approach, e.g.
python TPS.py -n 5
python TPS.py -n 10 --M_max 5 --M_min 4 --M_distr 0 --A_max 2000 --A_min 200 --A_distr 2
The full list of options is [default]: Depreciated (yet still functioning) parameters are given in {}.
--M1_max {--M_max} upper limit for the inner primary mass [100 Msun]
--M1_min {--M_min} lower limit for the inner primary mass [0.1 Msun]
--M1_distr {--M_distr} mass function option:
0: "Kroupa", #default
1: "Scalo",
2: "Miller & Scalo",
3: "Salpeter",
4: "Logarithmically flat",
5: "Eggleton",
6: "Kroupa for massive stars M>0.5 powerlaw with exp=-2.3",
--Qin_max {--Q_max} upper limit for the inner mass ratio [1.]
--Qin_min {--Q_min} lower limit for the inner mass ratio [0.]
--Qin_distr {--Q_distr} inner mass ratio option:
0: "Uniform distribution", #default
1: "Kroupa IMF",
2: "Galicher et al. 2016 powerlaw (M^-1.31)", #draws from mass distribution instead of mass ratio distribution,
--Qout_max {--q_max} upper limit for the outer mass ratio [1.]
--Qout_min {--q_min} lower limit for the mass of the outer star [0.]
--Qout_distr {--q_distr} outer mass ratio option:
0: "Uniform distribution", #default
1: "Kroupa IMF",
2: "Galicher et al. 2016 powerlaw (M^-1.31)", #draws from mass distribution instead of mass ratio distr
